A method and system for rapid breach situation awareness based on multi-source data

Abstract

The present invention relates to the technical field of breach image processing, and in particular to a method and system for rapid situational awareness of breaches based on multi-source data. The method comprises the following steps: acquiring dam monitoring station data; performing hydrological sensor network settings on the dam monitoring station data to obtain dam hydrological sensor network data; performing ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; performing breach node perception on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data; performing macro-disaster risk analysis of the breach area according to the breach node data to generate a macro-disaster risk schematic diagram of the breach area. The present invention improves the accuracy and response speed of rapid situational awareness of breaches through real-time data integration, comprehensive risk assessment, drone monitoring, dynamic prediction models and multi-party collaborative response.

Description

Technical Field

[0001] The present invention relates to the technical field of breach image processing, and in particular to a breach rapid situation awareness method and system based on multi-source data. Background Art

[0002] In the early days, situational awareness mainly relied on a single data source, such as satellite images or ground sensors, resulting in slow information acquisition and limited coverage. With the advancement of technology, data collection methods have gradually diversified, integrating multiple data sources such as remote sensing, drone monitoring, and social media information. Entering the 21st century, the rapid development of data processing and analysis technology has promoted the application of big data technology. In particular, the rise of machine learning and artificial intelligence has made real-time processing of complex data possible. Multi-source data fusion technology has emerged, which can integrate data of different types and dimensions, thereby improving the accuracy and timeliness of situational awareness. In recent years, with the popularization of the Internet of Things and 5G technology, the number of sensors and data transmission speed have increased significantly, making real-time monitoring and rapid response a reality. In addition, the introduction of blockchain technology provides a guarantee for the secure sharing of data and promotes cross-departmental and cross-regional information flow. However, at present, traditional breach situational awareness is often limited to local data, and the path for capturing situational awareness images of multiple breaches often lacks precise planning, resulting in low accuracy and responsiveness of rapid breach situational awareness. Summary of the invention

[0003] Based on this, it is necessary to provide a method and system for rapid situational awareness of a breach based on multi-source data to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a method for rapid breach situation awareness based on multi-source data is provided, the method comprising the following steps:

[0005] Step S1: Acquire the dam monitoring site data; perform hydrological sensor network setting on the dam monitoring site data to obtain the dam hydrological sensor network data; perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data;

[0006] Step S2: Perform macro-hazard risk analysis of the breach area according to breach node data, and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data, and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV according to the UAV flight route planning path data, and generate a micro-disaster diagram of dam damage; perform image integration of the macro-disaster diagram of the breach area and the micro-damage diagram of the dam area, and generate a rapid situation awareness image of the breach;

[0007] Step S3: extracting the breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; constructing a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map;

[0008] Step S4: setting a situation warning threshold for the breach situation prediction coding map to obtain a situation warning threshold; performing a situation warning on the breach rapid situation awareness image according to the situation warning threshold to generate a situation warning map; performing a multi-party collaborative situation awareness response on the situation warning map to generate a visual breach situation awareness shared atlas.

[0009] The present invention realizes real-time monitoring and ultrasonic water level change trend analysis by acquiring the data of the dam monitoring station and setting up the hydrological sensor network, effectively improving the early perception ability of the breach node, thereby providing an accurate basis for the subsequent risk assessment. By setting the situation warning threshold and conducting multi-party collaborative response, the situation warning map can be generated in time, promoting information sharing and cooperation between departments, greatly improving the efficiency and response speed of emergency management, and ensuring that a quick and effective response can be made in emergencies. Through the macro-disaster risk analysis of the breach node data and the UAV route planning, the disaster risk of the breach area can be comprehensively assessed, and the micro-dam damage analysis can be performed. This dual analysis improves the cognition of the overall and local damage of the dam and ensures a more comprehensive situation awareness. The macro-disaster risk analysis based on the breach node data can identify high-risk areas and generate schematic diagrams, providing decision makers with a clear risk distribution view. At the same time, through the UAV flight route planning, the latest data on the damage of the dam can be efficiently obtained. Spatial feature extraction of the rapid situation awareness image of the breach helps to quantify and analyze the specific characteristics of the breach. By constructing a situation prediction model, dynamic monitoring and prediction of breach situations can be achieved, so that decision makers can grasp future trends in a timely manner and take preventive measures. In addition, the generation of spatiotemporal evolution helps to reveal the key change points in the development of breaches, further improving the scientific nature of emergency management. The analysis of the degree of damage to the micro-dam, combined with macro and micro information, ensures the comprehensiveness of situation awareness and helps to better formulate emergency response plans. Extracting spatial features from the rapid situation awareness images of breaches and constructing a situation prediction model can help to achieve dynamic prediction of future situations, improve the foresight of emergency response, and ensure that relevant decisions are more scientifically based. By setting situation warning thresholds and conducting multi-party collaborative responses, situation warning maps can be generated in a timely manner, promoting information sharing and cooperation between departments, greatly improving the efficiency and response speed of emergency management, and ensuring that responses can be made quickly and effectively in emergencies. Setting situation warning thresholds enables potential crises to be discovered at an early stage and quickly communicated through situation warning maps, providing decision makers with an important time window. Multi-party collaborative situation awareness response promotes the efficient flow of information, ensures that relevant departments can share key information and coordinate response measures in a timely manner, and improves the efficiency and effectiveness of the overall emergency response. This collaborative mechanism is particularly important in emergencies, helping to reduce losses and improve the success rate of rescue. Therefore, the present invention improves the accuracy and response speed of rapid situational awareness of breaches through real-time data integration, comprehensive risk assessment, drone monitoring, dynamic prediction models and multi-party collaborative response.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire dam monitoring site data;

[0012] Step S12: performing hydrological sensor network setting on the dam monitoring station data to obtain dam hydrological sensor network data, wherein the dam hydrological sensor network includes an ultrasonic water level meter and a pressure sensor; performing ultrasonic water level measurement data collection on the dam hydrological sensor network data to obtain ultrasonic water level collection data;

[0013] Step S13: preprocessing the ultrasonic water level acquisition data to generate standard ultrasonic water level acquisition data, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization;

[0014] Step S14: Analyze the water level change trend of the standard ultrasonic water level acquisition data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data based on the ultrasonic water level change trend graph based on the pressure sensor to generate breach node data and breach prediction node data.

[0015] The present invention collects hydrological data near the dam in real time through a hydrological sensor network composed of ultrasonic water level meters and pressure sensors, covering various information such as water level and pressure, thereby improving the comprehensiveness and accuracy of monitoring data. Through data cleaning, missing value filling and standardization processing, the quality and consistency of the collected data are ensured, data noise and errors are reduced, and the accuracy of the analysis results is improved. The water level change trend chart is generated through ultrasonic water level measurement data, so that dam managers can monitor the changes in water levels in real time and detect abnormalities early. By combining pressure sensors with water level change trends, potential breach nodes can be sensed and predicted, which can provide early warning before the breach occurs, help take measures in advance, and prevent disasters. The generated breach node data and predicted node data can provide a reliable basis for subsequent dam maintenance and management decisions, and help managers make more accurate responses when the risk of breach increases.

[0016] Preferably, step S14 includes the following steps:

[0017] Step S141: performing timestamp and water level extraction on the standard ultrasonic water level acquisition data to obtain water level acquisition timestamp and water level data; using the water level acquisition timestamp as the x-axis and the water level data as the y-axis to perform time series discrete point mapping, thereby generating a water level change line graph;

[0018] Step S142: Calculate the water level change rate of the water level change line graph to generate water level change rate data; segment the water level change line graph into fluctuation change line segments according to the water level change rate data to generate an ultrasonic water level change trend graph, wherein the ultrasonic water level change trend graph includes fluctuation change line segments and gentle change line segments; perform abnormal fluctuation screening on the fluctuation change line segments based on a preset change rate threshold to obtain abnormal fluctuation line segment data;

[0019] Step S143: Perform trend line fitting on the gently changing line segments to generate gently changing trend line segments; perform abnormal fluctuation screening on the fluctuating line segments based on a preset change rate threshold to obtain abnormal fluctuation fitting line segment data; perform breach prediction area division on the dam hydrological sensor network data according to the gently changing trend line segments to generate breach first prediction area data; perform suspicious breach area division on the dam hydrological sensor network data according to the abnormal fluctuation line segment data to generate suspicious breach area data;

[0020] Step S144: Based on the pressure sensor, the dam area pressure analysis is performed on the suspected breach area data and the breach first prediction area data to generate the dam area pressure data; the breach node perception is performed on the dam hydrological sensor network data through the dam area pressure data to generate the breach node data and the breach prediction node data.

[0021] Through the extraction of timestamps and water level heights, the system can comprehensively consider time factors and water level changes, enhance the depth of data analysis, and generate a water level change line chart that provides managers with clear trend visualization, making it easy to identify violent fluctuations and stable states of water levels, and supporting quick decision-making. The calculation of the water level change rate enables the system to identify and respond to instantaneous changes, helping managers to understand the severity of potential risks in a timely manner. The abnormal fluctuation screening mechanism based on the preset change rate threshold can automatically identify abnormal water level changes, provide managers with accurate alarms, and reduce errors in human judgment. The fitting of the gently changing line segments provides data-driven prediction capabilities, enabling managers to foresee the trend of water level changes in advance and provide decision support for subsequent management. Through the analysis of the water level change line segments, the system can accurately divide the suspected breach area and the breach prediction area, which helps to concentrate resources for monitoring and maintenance of high-risk areas. Based on the data analysis of the pressure sensor, the pressure state of the dam at different water levels can be comprehensively evaluated, and the understanding of the overall health of the dam can be improved. Combined with the analysis of water level and pressure data, the system realizes multi-level breach node perception, and can quickly take early warning and response measures when the breach risk increases. Through data-driven methodology, managers can formulate scientific response strategies based on reliable analysis results and enhance their ability to respond to emergencies. The continuity of data collection and analysis provides a basis for system improvement and optimization, which can continuously improve the accuracy and efficiency of monitoring over time.

[0022] Preferably, the breach node perception of the dam hydrological sensor network data through the dam regional pressure data includes:

[0023] The pressure data of the dam area is used to detect the pressure. When the pressure data of the dam area is detected to be empty, the breach node of the dam hydrological sensor network data is marked to generate breach node data.

[0024] When the dam area pressure data detection is not empty, the dam area pressure data is screened for the dam double-side pressure to generate the first lateral pressure data and the second lateral pressure data; the first lateral pressure data and the second lateral pressure data are calculated for the pressure balance difference to obtain the dam pressure lateral difference data;

[0025] The suspected breach orientation is estimated through the lateral difference data of the dam pressure to obtain the breach second prediction area data; the breach nodes of the dam hydrological sensor network data are marked according to the breach first prediction area data and the breach second prediction area data to generate breach prediction node data.

[0026] The present invention can quickly mark potential breach nodes through pressure detection when the detection is empty, thereby improving the timeliness and accuracy of risk identification. By screening the pressure on both sides of the dam, balanced monitoring of both sides of the dam is ensured, reducing misjudgment caused by unilateral data distortion. Pressure balance difference calculation helps to deeply analyze the stress state of the dam, timely discover structural problems, and ensure the safety of the dam. Using lateral difference data to estimate the position of suspected breaches improves the accurate positioning of potential problem areas and contributes to more efficient resource allocation and management. Combined with the first and second prediction area data of the breach, the generated breach prediction node data provides a comprehensive risk warning and enhances the scientificity and foresight of decision-making. This method establishes a dynamic feedback mechanism for the dam monitoring system, which can respond to environmental changes in real time and improve the flexibility and adaptability of monitoring. Through accurate breach node perception, managers can rely on data to make more scientific countermeasures, reduce the risk of breaches, and ensure the safety of the dam. Through continuous data collection and analysis, the monitoring system can be continuously optimized to improve the efficiency and accuracy of overall monitoring.

[0027] Preferably, step S2 comprises the following steps:

[0028] Step S21: collecting satellite remote sensing images of the breach area according to breach node data to generate satellite remote sensing images of the breach area; performing macro-hazard risk analysis of the breach area on the satellite remote sensing images of the breach area to generate a macro-hazard risk diagram of the breach area;

[0029] Step S22: sorting the breach node data and the breach prediction node data by urgency, marking the breach node data as high-urgency node data, and marking the breach prediction node data as low-urgency node data; performing UAV flight route planning based on the high-urgency node data and the low-urgency node data, and generating UAV flight route planning path data;

[0030] Step S23: taking microscopic high-precision breach photos according to the UAV flight route planning path data, thereby generating a microscopic high-precision breach image; analyzing the degree of dam damage on the microscopic high-precision breach image, and generating a microscopic disaster diagram of dam damage;

[0031] Step S24: integrating the macroscopic disaster diagram of the breach area and the microscopic damage diagram of the dam area to generate a rapid situation awareness image of the breach.

[0032] The present invention can quickly obtain a macroscopic view of the breach area through the collection of satellite remote sensing images, which is helpful to fully understand the scope and characteristics of disaster impact. The satellite remote sensing images are subjected to macroscopic disaster risk analysis, and the generated schematic diagram provides an intuitive representation of potential risks, supporting decision makers to identify and assess risks. The breach nodes and predicted node data are sorted by urgency, so that managers can prioritize nodes with high urgency and optimize resource allocation and response strategies. The flight route of the unmanned aerial vehicle is planned based on the urgency data to ensure that the flight mission of the unmanned aerial vehicle is efficient and directly reaches the key areas, thereby improving the pertinence of monitoring. The microscopic high-precision photography of the unmanned aerial vehicle can record the specific situation of the damage to the dam in detail, provide important detailed information, and support subsequent evaluation and repair. The microscopic high-precision image is used to analyze the degree of damage to the dam, and the generated microscopic disaster schematic diagram provides a reliable basis for damage assessment and helps decision makers to formulate a repair plan. The macroscopic disaster schematic diagram is integrated with the microscopic damage schematic diagram, and the generated rapid situation awareness image provides a global perspective, helping managers to make quick judgments in complex situations. By integrating data and information at different levels, comprehensive situational awareness is provided, scientific decision-making is supported, and the ability to respond to emergencies is improved. The implementation of this step provides a dynamic feedback mechanism for risk management, which can continuously optimize monitoring and response strategies based on actual conditions.

[0033] Preferably, performing macro-hazard risk analysis of the breach area on the satellite remote sensing image of the breach area includes:

[0034] Performing connected time-series image screening on the satellite remote sensing images of the breach area to obtain a first satellite remote sensing time-series image of the breach area, a second satellite remote sensing time-series image of the breach area, and a third satellite remote sensing time-series image of the breach area, wherein the first satellite remote sensing time-series image of the breach area is a previous frame image of the second satellite remote sensing time-series image of the breach area, and the third satellite remote sensing time-series image of the breach area is a subsequent frame image of the second satellite remote sensing time-series image of the breach area;

[0035] Performing image pixel value difference calculation on the first satellite remote sensing time series image of the breach area and the second satellite remote sensing time series image of the breach area to obtain a first image pixel difference value; performing image pixel value difference calculation on the second satellite remote sensing time series image of the breach area and the third satellite remote sensing time series image of the breach area to obtain a second image pixel difference value;

[0036] According to the first image pixel difference value and the second image pixel difference value, the satellite remote sensing image of the breach area is segmented into a change area to obtain a satellite remote sensing change area image of the breach area; the first image pixel difference value and the second image pixel difference value are analyzed for positive and negative changes to generate image pixel positive difference values ​​and image pixel negative difference values;

[0037] Based on the positive difference of image pixels, the satellite remote sensing change area image of the breach area is analyzed for terrain elevation changes to generate terrain elevation change data; based on the negative difference of image pixels, the satellite remote sensing change area image of the breach area is used to detect water body changes to generate water body change detection data; the terrain elevation change data and water body change detection data are used to visualize the macro-disaster risk of the breach area to generate a macro-disaster risk schematic diagram of the breach area.

[0038] The present invention ensures that the monitoring of the breach area is not only static, but also captures dynamic changes through the screening of connected time series images, thereby enhancing the timeliness of the analysis. Image pixel value difference calculation can accurately identify the change area in the time series, providing a solid data basis for subsequent disaster assessment. Through the segmentation of the change area, the system can effectively extract the affected area, reduce the interference of irrelevant information, and make the analysis more focused and accurate. Positive and negative change analysis can not only reveal the changes in terrain, but also effectively detect the changes in water bodies, providing a multi-dimensional perspective for the comprehensive assessment of disaster risks. The terrain elevation change analysis based on the positive difference of image pixels can provide quantitative data on terrain changes, which helps to understand the potential impact of terrain changes on dam safety. The implementation of water body change detection enables managers to grasp the dynamics of water bodies in a timely manner and identify water level changes or water body expansion that lead to breach risks. The terrain elevation change data and water body change detection data are visualized, and the generated macro-disaster hazard diagram can intuitively display the spatial distribution of disaster risks and support decision makers to respond effectively. By integrating time series changes and multi-dimensional data, more comprehensive disaster assessment information is provided to managers, which helps to formulate scientific disaster prevention and mitigation strategies. This analysis step establishes a dynamic feedback mechanism for the disaster monitoring system, which can continuously optimize monitoring and response strategies based on real-time data. Through comprehensive and accurate macro-disaster risk analysis, managers can respond quickly when disasters occur, reduce potential losses, and improve the efficiency of emergency management.

[0039] Preferably, performing flight path planning of the UAV according to the high-urgency node data and the low-urgency node data includes:

[0040] Perform node connection on high-urgency node data and low-urgency node data to generate a node connection network; perform priority path series connection on high-urgency node data through the node connection network to generate a high-urgency node flight path;

[0041] Confirm the node position of the high-urgency node flight path to obtain the high-urgency node position data; search for low-urgency nodes on the high-urgency node flight path through the high-urgency node position data to generate low-urgency node data with the same position;

[0042] The same-direction low-urgency node data and high-urgency node flight paths are merged into the same-direction paths to generate the same-direction node flight paths; the node connection network is segmented according to the same-direction node flight paths to generate the same-direction UAV flight paths; the adjacent channel angles are calculated for the same-direction UAV flight paths to obtain the adjacent channel angle data;

[0043] The flight course of the UAV with the same azimuth is transformed into the flight course with the minimum angle through the angle data of adjacent courses, and the flight course planning path data of the UAV is generated.

[0044] The present invention generates a flight path for high-emergency nodes by connecting high-emergency nodes in series with priority paths, ensuring that the UAV arrives at the area that needs emergency treatment most first, thereby improving the efficiency and accuracy of emergency response. Node position confirmation and the merging of the same-orientation paths of low-emergency nodes reduce unnecessary path duplication and redundant flights of the UAV during flight, optimize flight time and power consumption, and improve the efficiency of UAV flight. According to the same-orientation path merging strategy of high-emergency and low-emergency nodes, it can dynamically adapt to different situations to ensure that all key areas can be covered while avoiding the limitations of a single priority node processing strategy. Flight path segmentation allows the UAV to fly in segments in complex terrain or mission areas, ensuring that each section of the route is reasonably planned according to actual needs and node urgency, making task execution more flexible. By calculating the angle between adjacent routes and transforming the minimum angle, the number of flight turns of the UAV can be significantly reduced, reducing the energy consumption and flight time extension caused by turning, making the UAV path smoother and more efficient. The path merging of low-urgency nodes in the same direction optimizes the overall mission planning, avoids redundant single flight mission arrangements, saves valuable flight time and power resources, and improves the operational capabilities of the drone group. During the flight, based on real-time azimuth data and path angle information, the drone can adaptively adjust the flight path to ensure the flexibility and real-time nature of the mission, which helps to cope with sudden changes. By changing the flight path in the minimum angle azimuth, the drone can avoid frequent sharp turns, fly more stably, and reduce the risk of flight failures. The node connection network and channel segmentation technology can provide accurate channel division for multiple drones, avoid channel conflicts and repeated flights, and improve the efficiency and safety of overall air traffic management. This process can be applied to node networks of various sizes and urgency level distributions, and can be expanded to a larger range of mission areas by adjusting node priorities and flight path planning, providing support for complex drone formation flights.

[0045] Preferably, analyzing the damage degree of the dam using the microscopic high-precision breach image includes:

[0046] The breach area contour is segmented on the microscopic high-precision breach image to generate the breach area contour image; the number of pixels in the damaged area is calculated on the breach area contour image to obtain the number of pixels in the damaged area; the number of pixels in the damaged area and the microscopic high-precision breach image are converted into actual geographic size ratios to generate the dam damaged area data;

[0047] Perform multi-view image screening on microscopic high-precision breach images to generate a multi-view image set of microscopic breaches; perform feature point extraction on the multi-view image set of microscopic breaches to generate multi-view image feature points; perform feature point matching on the multi-view image feature points to generate multi-view matching feature points;

[0048] Parallax calculation is performed on multi-view matching feature points to obtain the multi-view parallax calculation results of the breach; binocular reconstruction is performed on the multi-view parallax calculation results of the breach to generate a dam damage depth map; depth estimation is performed on the dam damage depth map to generate dam damage depth data; the dam damage area data and dam damage depth data are visualized to generate a macro-disaster hazard diagram of the breach area.

[0049] The present invention can effectively extract the specific shape of the damaged area through the segmentation of the breach area contour, improve the accuracy of subsequent analysis, and reduce misjudgment. The calculation of the number of pixels in the damaged area is combined with the conversion of the geographic size ratio to provide quantitative data on the damaged area of ​​the dam, providing an important basis for post-disaster assessment. The generation of a multi-view image set of micro breaches allows the breach to be examined from different angles, enriches the understanding of the damage situation, and reduces the information loss caused by the limitation of the viewing angle. The accurate correspondence between the multi-view images is ensured by feature point extraction and matching, making the subsequent parallax calculation and reconstruction results more reliable. Parallax calculation provides the necessary data support for the depth analysis of the dam, so that the assessment of the damage is not limited to the surface, but can also go deep into the structural level. The dam damage depth map generated by binocular reconstruction technology provides detailed depth information for disaster response and repair work, helping relevant personnel to formulate repair strategies. The damaged area data and depth data are visualized, and the generated macro-disaster hazard schematic diagram intuitively shows the overall damage of the dam, which is convenient for decision makers to quickly understand and respond. Detailed data on the area and depth of damage provide a scientific basis for dam safety management and subsequent repair plans, helping to rationally allocate resources and formulate effective emergency measures. This analysis process improves the monitoring and assessment capabilities of dam damage, allowing for rapid and comprehensive assessments after disasters occur, reducing potential risks.

[0050] Preferably, step S3 comprises the following steps:

[0051] Step S31: extracting the breach space features from the breach rapid situation awareness image to obtain a breach space feature matrix, wherein the breach space feature extraction includes texture feature extraction and shape feature extraction;

[0052] Step S32: dividing the breach space feature matrix into a data set to obtain a model training set and a model test set; training the model training set using a long short-term memory neural network algorithm to generate a breach situation prediction pre-model; optimizing and iterating the breach situation prediction pre-model using the model test set to generate a breach situation prediction model;

[0053] Step S33: importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction map;

[0054] Step S34: color-code the breach situation prediction map to generate a breach situation prediction coded map.

[0055] The present invention extracts texture and shape features to generate a breach spatial feature matrix that can fully reflect the structure and properties of the breach, laying the foundation for subsequent analysis. The data set is divided into a model training set and a test set to ensure the generalization ability of the model on different data, reduce overfitting, and thus improve the accuracy of the prediction. The long short-term memory neural network (LSTM) is used for model training to adapt to the characteristics of time series data, so that the model can capture the temporal relationship of the change in the breach situation. The prediction model is optimized and iterated through the model test set to ensure continuous improvement of the model and improve the prediction ability and accuracy of the future breach situation. The rapid situation awareness image of the breach is imported into the prediction model, which can monitor and analyze the spatiotemporal evolution trend of the breach in real time, providing data support for rapid response and decision-making. The breach situation prediction map is visualized through color coding to intuitively display the development trend of the breach, so that decision makers can quickly understand the current situation and take necessary measures. Combined with spatial features and spatiotemporal evolution analysis, the prediction model can provide accurate information support for the disaster response team and improve the efficiency and effectiveness of the overall emergency response. The breach situation prediction model can be used for early warning to help relevant departments take necessary preventive measures before potential risks occur. Through accurate prediction maps, decision makers can make more scientific and reasonable decisions in complex situations and reduce the impact of disasters. The generated situation prediction coding maps can be shared with relevant agencies to promote cooperation among multiple departments and improve overall disaster management efficiency.

[0056] In this specification, a breach rapid situation awareness system based on multi-source data is provided, which is used to execute the above-mentioned breach rapid situation awareness method based on multi-source data. The breach rapid situation awareness system based on multi-source data includes:

[0057] The breach sensing module is used to obtain the data of the dam monitoring station; set up the hydrological sensor network for the dam monitoring station data to obtain the dam hydrological sensor network data; perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; perform breach node sensing on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data;

[0058] The breach image analysis module is used to perform macro-hazard risk analysis of the breach area based on breach node data and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV flight route planning path data and generate a micro-hazard diagram of dam damage; perform image integration of the macro-hazard diagram of the breach area and the micro-damage diagram of the dam area to generate a rapid situation awareness image of the breach;

[0059] The breach dynamic evolution module is used to extract breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; construct a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; import the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map;

[0060] The breach perception and warning module is used to set the situation warning threshold for the breach situation prediction coding map to obtain the situation warning threshold; perform situation warning on the breach rapid situation perception image according to the situation warning threshold to generate a situation warning map; perform multi-party collaborative situation perception response on the situation warning map to generate a visual breach situation perception shared atlas.

[0061] The beneficial effect of the present invention is that by acquiring the data of the dam monitoring station and setting up the hydrological sensor network, the collected data can be ensured to be accurate and comprehensive, laying the foundation for subsequent analysis. Through ultrasonic water level change trend analysis, water level changes can be monitored in real time, abnormal conditions can be discovered in time, and it is helpful to respond quickly to potential risks. The breach node data and prediction node data generated according to the water level change trend are helpful to identify and predict breach risks and improve the predictability of emergency management. The macro-disaster risk analysis of the breach area and the generated danger diagram can provide decision makers with a clear risk view and help evaluate potential impacts. The UAV flight route planning based on the breach node data can optimize the UAV monitoring path, improve monitoring efficiency, and ensure that key areas receive focus. The micro-dam damage degree analysis can provide a detailed damage assessment, help decision makers understand the actual damage situation, and formulate corresponding repair plans. The macro and micro damage diagrams are image-integrated, and the generated rapid situation awareness image provides decision makers with an intuitive perspective and helps to quickly judge the current situation. The feature matrix obtained by spatial feature extraction can reveal the spatial distribution and characteristics of the breach and provide rich information for subsequent model construction. The situation prediction model constructed based on the feature matrix can make scientific predictions using historical data and current conditions, and enhance the foresight of disaster management. By importing the rapid situation awareness image into the prediction model for spatiotemporal evolution, the current situation can be dynamically updated to help decision makers adjust response strategies in a timely manner. By setting the situation warning threshold, effective monitoring of potential dangers can be achieved to ensure that appropriate measures are taken before the danger occurs. The generation of situation warning maps promotes the collaboration of all parties, ensures timely communication and collaboration between relevant departments, and improves the efficiency of emergency response. The generation of a shared atlas of visual breach situation awareness promotes the transparency and sharing of information, and provides strong decision-making support for managers at all levels. Therefore, the present invention improves the accuracy and response speed of rapid situation awareness of breaches through real-time data integration, comprehensive risk assessment, drone monitoring, dynamic prediction models, and multi-party collaborative response. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 A schematic diagram of the process flow of a breach rapid situation awareness method based on multi-source data;

[0063] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0064] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0065] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0066] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0067] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0068] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0069] To achieve this, please refer to Figures 1 to 3 A method for rapid situational awareness of a breach based on multi-source data, the method comprising the following steps:

[0070] Step S1: Acquire the dam monitoring site data; perform hydrological sensor network setting on the dam monitoring site data to obtain the dam hydrological sensor network data; perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data;

[0071] Step S2: Perform macro-hazard risk analysis of the breach area according to breach node data, and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data, and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV according to the UAV flight route planning path data, and generate a micro-disaster diagram of dam damage; perform image integration of the macro-disaster diagram of the breach area and the micro-damage diagram of the dam area, and generate a rapid situation awareness image of the breach;

[0072] Step S3: extracting the breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; constructing a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map;

[0073] Step S4: setting a situation warning threshold for the breach situation prediction coding map to obtain a situation warning threshold; performing a situation warning on the breach rapid situation awareness image according to the situation warning threshold to generate a situation warning map; performing a multi-party collaborative situation awareness response on the situation warning map to generate a visual breach situation awareness shared atlas.

[0074] The present invention realizes real-time monitoring and ultrasonic water level change trend analysis by acquiring the data of the dam monitoring station and setting up the hydrological sensor network, effectively improving the early perception ability of the breach node, thereby providing an accurate basis for the subsequent risk assessment. By setting the situation warning threshold and conducting multi-party collaborative response, the situation warning map can be generated in time, promoting information sharing and cooperation between departments, greatly improving the efficiency and response speed of emergency management, and ensuring that a quick and effective response can be made in emergencies. Through the macro-disaster risk analysis of the breach node data and the UAV route planning, the disaster risk of the breach area can be comprehensively assessed, and the micro-dam damage analysis can be performed. This dual analysis improves the cognition of the overall and local damage of the dam and ensures a more comprehensive situation awareness. The macro-disaster risk analysis based on the breach node data can identify high-risk areas and generate schematic diagrams, providing decision makers with a clear risk distribution view. At the same time, through the UAV flight route planning, the latest data on the damage of the dam can be efficiently obtained. Spatial feature extraction of the rapid situation awareness image of the breach helps to quantify and analyze the specific characteristics of the breach. By constructing a situation prediction model, dynamic monitoring and prediction of breach situations can be achieved, so that decision makers can grasp future trends in a timely manner and take preventive measures. In addition, the generation of spatiotemporal evolution helps to reveal the key change points in the development of breaches, further improving the scientific nature of emergency management. The analysis of the degree of damage to the micro-dam, combined with macro and micro information, ensures the comprehensiveness of situation awareness and helps to better formulate emergency response plans. Extracting spatial features from the rapid situation awareness images of breaches and constructing a situation prediction model can help to achieve dynamic prediction of future situations, improve the foresight of emergency response, and ensure that relevant decisions are more scientifically based. By setting situation warning thresholds and conducting multi-party collaborative responses, situation warning maps can be generated in a timely manner, promoting information sharing and cooperation between departments, greatly improving the efficiency and response speed of emergency management, and ensuring that responses can be made quickly and effectively in emergencies. Setting situation warning thresholds enables potential crises to be discovered at an early stage and quickly communicated through situation warning maps, providing decision makers with an important time window. Multi-party collaborative situation awareness response promotes the efficient flow of information, ensures that relevant departments can share key information and coordinate response measures in a timely manner, and improves the efficiency and effectiveness of the overall emergency response. This collaborative mechanism is particularly important in emergencies, helping to reduce losses and improve the success rate of rescue. Therefore, the present invention improves the accuracy and response speed of rapid situational awareness of breaches through real-time data integration, comprehensive risk assessment, drone monitoring, dynamic prediction models and multi-party collaborative response.

[0075] In the embodiment of the present invention, reference Figure 1The above is a schematic diagram of the steps of a method for rapid situational awareness of a breach based on multi-source data of the present invention. In this example, the method for rapid situational awareness of a breach based on multi-source data includes the following steps:

[0076] Step S1: Acquire the dam monitoring site data; perform hydrological sensor network setting on the dam monitoring site data to obtain the dam hydrological sensor network data; perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data;

[0077] In an embodiment of the present invention, monitoring stations are arranged at key locations of the dam, and hydrological sensors (such as water level meters, flow meters, etc.) are used to collect data. The recorded data include water level, flow rate, temperature, precipitation, etc., to ensure data integrity and accuracy. The collected monitoring data is uploaded to a cloud database or a local server for classified storage for subsequent analysis. Ultrasonic water level sensors are arranged at different locations of the dam (such as upstream, midstream, and downstream) to monitor water level changes in real time. The sensor spacing is determined to ensure coverage of the entire dam area to form an effective hydrological sensor network. Each sensor is connected through a wireless network (such as LoRa, NB-IoT, etc.) so that it can transmit data to the central monitoring system in real time. The network transmission protocol is configured to ensure stable data transmission and low latency. A data set of the dam hydrological sensor network is generated through sensor configuration, including the location, type, status, and real-time hydrological data of each sensor. The hydrological data transmitted to the server is cleaned to remove outliers and noise (such as sensor failure, weather influence, etc.). The data of different sensors are integrated into a unified data set, and the timestamps are aligned to ensure that each data can be analyzed in time series. Use time series analysis methods (such as moving average, weighted average, etc.) to calculate the water level change trend of each sensor in different time periods. Visualize the water level change data and generate an ultrasonic water level change trend graph to show the change of water level over time and help determine the trend of water level rise or fall. Set water level thresholds based on historical water level data and dam design standards to determine breach risk points (such as water level exceeding a certain critical value). Analyze the ultrasonic water level change trend graph to identify water level change nodes that exceed the threshold and mark them as breach nodes. Record the time, location and water level data of the breach node. Based on the breach node data, use prediction models (such as time series prediction, machine learning models, etc.) to generate prediction node data for future breaches. The prediction node should include the expected time, location and related water level change information.

[0078] Step S2: Perform macro-hazard risk analysis of the breach area according to breach node data, and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data, and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV according to the UAV flight route planning path data, and generate a micro-disaster diagram of dam damage; perform image integration of the macro-disaster diagram of the breach area and the micro-damage diagram of the dam area, and generate a rapid situation awareness image of the breach;

[0079] In an embodiment of the present invention, high-resolution image data of the breach area is obtained by satellite remote sensing technology based on breach node data. Connected time-series images are selected from the acquired satellite remote sensing images to generate remote sensing images of multiple time points (such as the previous frame, the current frame, and the next frame). Pixel value difference calculation is performed on the connected time-series images to identify the change area, generate a change area segmentation map, and identify the existing damaged or changed areas. The terrain elevation change analysis is performed on the change area to generate terrain elevation change data. The water body change is detected to generate water body change data. The terrain elevation change data and the water body change data are combined and visualized to generate a macroscopic disaster hazard diagram of the breach area. According to the breach node data and the breach prediction node data, the urgency is sorted, and the high urgency nodes are marked as high priority, and the low urgency nodes are marked as low priority. The high urgency nodes and the low urgency nodes are connected to form a node connection network. In the node connection network, the high urgency nodes are preferentially connected in series to generate a high priority flight path. The high priority flight path is combined with the low urgency nodes in the same direction to generate the flight path planning path data of the unmanned aerial vehicle. According to the planned flight path of the drone, high-precision micro breach photography is performed to obtain high-definition images of the dam. The contour of the micro high-precision breach image is segmented to generate a contour image of the breach area. The number of pixels in the damaged area is calculated to estimate the damaged area of ​​the dam. Feature point extraction and parallax calculation are performed on the multi-view images to generate a dam damage depth map. Combine the damaged area data and depth data to generate a micro-disaster schematic diagram of the damaged dam. The generated macro-disaster schematic diagram of the breach area is integrated with the micro-damage schematic diagram of the dam area, and a specific image processing algorithm (such as weighted averaging, transparency adjustment, etc.) is selected to achieve fusion. The integrated image is visually adjusted to generate the final breach rapid situation awareness image, which displays comprehensive macro and micro disaster risk information.

[0080] Step S3: extracting the breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; constructing a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map;

[0081] In an embodiment of the present invention, a rapid situation awareness image of a breach is imported into a feature extraction system. Texture analysis algorithms (such as gray level co-occurrence matrix, LBP, etc.) are used to extract texture features from the image to generate a texture feature matrix. Edge detection and contour extraction algorithms (such as Canny edge detection) are used to identify shape features to generate a shape feature matrix. The texture feature matrix is ​​fused with the shape feature matrix to obtain a comprehensive breach space feature matrix, covering a variety of feature information. The breach space feature matrix is ​​divided into a training set and a test set, usually 70% for training and 30% for testing. A specific prediction model (such as long short-term memory network LSTM, convolutional neural network CNN, etc.) is selected to construct a breach situation prediction model. The prediction model is trained using the training set, and the model parameters are optimized by the back propagation algorithm to improve the prediction accuracy. The model is verified and optimized using the test set, and the model structure or parameters are adjusted according to the prediction accuracy to ensure the generalization ability of the model. The rapid situation awareness image of the breach is imported into the constructed breach situation prediction model. The model is used to perform spatiotemporal evolution on the input image, analyze future situation changes, and generate situation prediction results. The generated situation prediction results are color-coded, different states or risk levels are distinguished by colors, and a breach situation prediction coding map is generated.

[0082] Step S4: setting a situation warning threshold for the breach situation prediction coding map to obtain a situation warning threshold; performing a situation warning on the breach rapid situation awareness image according to the situation warning threshold to generate a situation warning map; performing a multi-party collaborative situation awareness response on the situation warning map to generate a visual breach situation awareness shared atlas.

[0083] In an embodiment of the present invention, by analyzing the historical breach event data and the output results of the situation prediction model, key indicators (such as water level change rate, pressure change, etc.) are determined as the basis for early warning. According to the analysis results, multiple early warning thresholds are set, including warning values, warning values, and crisis values. These thresholds can be determined based on statistical methods (such as percentiles) or empirical rules. The set thresholds are backtested and their effectiveness is verified using historical data to ensure effective early warning in actual events. The situation warning threshold is applied to the breach rapid situation awareness image, and by comparing the real-time monitoring data (such as water level, pressure, etc.) with the threshold, it is determined whether the warning is triggered. If the monitoring data exceeds the set warning threshold, a situation warning map is generated, and the affected area and warning level (such as normal, warning, warning, crisis) are marked in the map. According to the situation warning map, the corresponding emergency response mechanism is activated to notify relevant departments and personnel. A multi-party collaborative response platform is established to integrate information from all parties (such as emergency management, scientific research institutions, etc.) to ensure information sharing and real-time updating. The situation warning map is integrated with relevant information (such as real-time monitoring data, resource allocation, emergency measures, etc.) to generate a visual breach situation awareness shared atlas. The atlas can contain information from different perspectives and dimensions to facilitate understanding and response by different departments and personnel.

[0084] Preferably, step S1 comprises the following steps:

[0085] Step S11: Acquire dam monitoring site data;

[0086] Step S12: performing hydrological sensor network setting on the dam monitoring station data to obtain dam hydrological sensor network data, wherein the dam hydrological sensor network includes an ultrasonic water level meter and a pressure sensor; performing ultrasonic water level measurement data collection on the dam hydrological sensor network data to obtain ultrasonic water level collection data;

[0087] Step S13: preprocessing the ultrasonic water level acquisition data to generate standard ultrasonic water level acquisition data, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization;

[0088] Step S14: Analyze the water level change trend of the standard ultrasonic water level acquisition data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data based on the ultrasonic water level change trend graph based on the pressure sensor to generate breach node data and breach prediction node data.

[0089] In the embodiment of the present invention, ultrasonic water level meters (e.g., working range, measurement accuracy) and pressure sensors (e.g., corrosion resistance) are selected. Sensors are arranged reasonably according to the topographic and hydrological characteristics of the dam. Ensure that there are sensors in key areas such as the middle section, both ends, and potential breach locations of the dam. Use fixed brackets or mounting platforms to ensure the stability of the sensors. Ensure that all sensors are connected to the central data processing system via wireless or wired networks, and configure data transmission protocols. Set thresholds to identify and remove abnormal values ​​(e.g., instantaneous high water levels) that exceed the normal range. Interpolation methods (e.g., linear interpolation or spline interpolation) are used to fill missing data and maintain data continuity. Water level data is converted into a unified unit (e.g., meter) to ensure consistency with historical data. Standard ultrasonic water level data is smoothed using moving average methods, exponential smoothing methods, etc., to generate an ultrasonic water level change trend graph. Analyze periodic fluctuations, mutation points, and long-term change trends in the trend graph. Use pressure sensors to monitor pressure changes at the bottom and sides of the dam, and identify potential breach areas in combination with ultrasonic water level change data. Use machine learning models (such as support vector machines or random forests) to train data to improve the accuracy of breach node prediction. Aggregate breach node data and breach prediction node data to generate visual reports and provide decision support.

[0090] Preferably, step S14 includes the following steps:

[0091] Step S141: performing timestamp and water level extraction on the standard ultrasonic water level acquisition data to obtain water level acquisition timestamp and water level data; using the water level acquisition timestamp as the x-axis and the water level data as the y-axis to perform time series discrete point mapping, thereby generating a water level change line graph;

[0092] Step S142: Calculate the water level change rate of the water level change line graph to generate water level change rate data; segment the water level change line graph into fluctuation change line segments according to the water level change rate data to generate an ultrasonic water level change trend graph, wherein the ultrasonic water level change trend graph includes fluctuation change line segments and gentle change line segments; perform abnormal fluctuation screening on the fluctuation change line segments based on a preset change rate threshold to obtain abnormal fluctuation line segment data;

[0093] Step S143: Perform trend line fitting on the gently changing line segments to generate gently changing trend line segments; perform abnormal fluctuation screening on the fluctuating line segments based on a preset change rate threshold to obtain abnormal fluctuation fitting line segment data; perform breach prediction area division on the dam hydrological sensor network data according to the gently changing trend line segments to generate breach first prediction area data; perform suspicious breach area division on the dam hydrological sensor network data according to the abnormal fluctuation line segment data to generate suspicious breach area data;

[0094] Step S144: Based on the pressure sensor, the dam area pressure analysis is performed on the suspected breach area data and the breach first prediction area data to generate the dam area pressure data; the breach node perception is performed on the dam hydrological sensor network data through the dam area pressure data to generate the breach node data and the breach prediction node data.

[0095] In an embodiment of the present invention, the timestamp and the corresponding water level height of each sampling point are extracted from the standard ultrasonic water level acquisition data. The water level acquisition timestamp is used as the x-axis and the water level height data is used as the y-axis to generate a time series line graph, which is convenient for observing the change trend of the water level. According to the water level height difference between adjacent timestamps, the water level change rate (Δ water level / Δ time) is calculated to generate water level change rate data. According to the water level change rate data, the water level change line graph is segmented to identify the fluctuating change line segments and the gently changing line segments. A preset change rate threshold is set, and the fluctuating change line segments exceeding the threshold are screened out to obtain abnormal fluctuating line segment data. Linear regression or polynomial fitting is used for the gently changing line segments to generate gently changing trend line segments. Regression analysis is performed on the screened abnormal fluctuation line segments to generate abnormal fluctuation fitting line segment data. According to the gently changing trend line segments, the dam hydrological sensor network data is analyzed to divide the first prediction area data of the breach. According to the abnormal fluctuation line segment data, the suspicious breach area is identified and the suspicious breach area data is generated. Using the pressure data monitored by the pressure sensor, the suspected breach area and the first predicted breach area are analyzed to generate the pressure data of the dam area. By analyzing the correlation between the pressure data of the dam area and the water level change data, the breach nodes are identified, and the breach node data and breach prediction node data are generated. Data mining technology (such as cluster analysis or anomaly detection algorithm) is used to improve the recognition accuracy of breach nodes.

[0096] Preferably, the breach node perception of the dam hydrological sensor network data using the dam regional pressure data includes:

[0097] The pressure data of the dam area is used to detect the pressure. When the pressure data of the dam area is detected to be empty, the breach node of the dam hydrological sensor network data is marked to generate breach node data.

[0098] When the dam area pressure data detection is not empty, the dam area pressure data is screened for the dam double-side pressure to generate the first lateral pressure data and the second lateral pressure data; the first lateral pressure data and the second lateral pressure data are calculated for the pressure balance difference to obtain the dam pressure lateral difference data;

[0099] The suspected breach orientation is estimated through the lateral difference data of the dam pressure to obtain the breach second prediction area data; the breach nodes of the dam hydrological sensor network data are marked according to the breach first prediction area data and the breach second prediction area data to generate breach prediction node data.

[0100] In an embodiment of the present invention, the validity of the pressure data in the dam area is checked. If the data is detected to be empty, it indicates that there is a potential breach risk. When the pressure data is empty, the dam hydrological sensor network data is analyzed, the corresponding breach node is marked, and the breach node data is generated. If the dam area pressure data is not empty, the pressure analysis is continued. The dam area pressure data is screened, and the first lateral pressure data and the second lateral pressure data are extracted respectively. According to the first lateral pressure data and the second lateral pressure data, the pressure balance difference is calculated to obtain the dam pressure lateral difference data (the calculation formula is: ΔP = P1 - P2, where P1 and P2 are the first lateral and second lateral pressure data, respectively). The dam pressure lateral difference data is evaluated to determine the degree of pressure imbalance. The dam pressure lateral difference data is used to estimate the suspected breach orientation, and the breach direction is determined in combination with the pressure difference, and the breach second prediction area data is generated. Based on the breach first prediction area data and the breach second prediction area data, the dam hydrological sensor network data is further analyzed, the breach nodes are marked, and the breach prediction node data is generated.

[0101] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0102] Step S21: collecting satellite remote sensing images of the breach area according to breach node data to generate satellite remote sensing images of the breach area; performing macro-hazard risk analysis of the breach area on the satellite remote sensing images of the breach area to generate a macro-hazard risk diagram of the breach area;

[0103] Step S22: sorting the breach node data and the breach prediction node data by urgency, marking the breach node data as high-urgency node data, and marking the breach prediction node data as low-urgency node data; performing UAV flight route planning based on the high-urgency node data and the low-urgency node data, and generating UAV flight route planning path data;

[0104] Step S23: taking microscopic high-precision breach photos according to the UAV flight route planning path data, thereby generating a microscopic high-precision breach image; analyzing the degree of dam damage on the microscopic high-precision breach image, and generating a microscopic disaster diagram of dam damage;

[0105] Step S24: integrating the macroscopic disaster diagram of the breach area and the microscopic damage diagram of the dam area to generate a rapid situation awareness image of the breach.

[0106] In an embodiment of the present invention, image acquisition is performed by using a high-resolution satellite. Data analysis is performed using GIS (Geographic Information System) software. The coordinates of the breach area are determined, and a satellite remote sensing platform (such as Sentinel-2 or WorldView) is used to acquire images. A scheduled task is set to regularly acquire image data to ensure dynamic monitoring. The acquired images are imported into the GIS software for image preprocessing (such as radiation correction and geometric correction). Potential risk areas are identified by classification algorithms (such as supervised classification and unsupervised classification), a macroscopic disaster hazard diagram is generated, and high-risk areas are identified. Using data analysis tools, relevant information (such as risk level and historical data) of breach nodes and prediction nodes is input into the model. A sorting algorithm (such as a weighted scoring method) is used to sort the nodes by urgency. Based on high and low urgency nodes, a route planning is performed using drone aerial photography software (such as DJI GroundStation or Mission Planner). Route parameters are set to ensure that the drone can cover all high-urgency areas. According to the planned route, the drone is dispatched for microscopic high-precision photography. Ensure that the shooting parameters (such as exposure and ISO) are optimized to obtain clear images. Import the captured images into image processing software (such as MATLAB, OpenCV). Use image analysis algorithms (such as edge detection and texture analysis) to assess the damage to the dam and generate a microscopic disaster diagram of the damage, marking the damaged area and its extent. Use image processing software to import macro and micro images, and perform stitching and overlay. Apply image fusion algorithms to improve the readability and information content of the overall image. Overlay the integrated images in the GIS software, and add legends and explanatory information. Output the final breach rapid situation awareness image to ensure that it is suitable for on-site command and decision support.

[0107] Preferably, performing macro-hazard risk analysis of the breach area on the satellite remote sensing image of the breach area includes:

[0108] Performing connected time-series image screening on the satellite remote sensing images of the breach area to obtain a first satellite remote sensing time-series image of the breach area, a second satellite remote sensing time-series image of the breach area, and a third satellite remote sensing time-series image of the breach area, wherein the first satellite remote sensing time-series image of the breach area is a previous frame image of the second satellite remote sensing time-series image of the breach area, and the third satellite remote sensing time-series image of the breach area is a subsequent frame image of the second satellite remote sensing time-series image of the breach area;

[0109] Performing image pixel value difference calculation on the first satellite remote sensing time series image of the breach area and the second satellite remote sensing time series image of the breach area to obtain a first image pixel difference value; performing image pixel value difference calculation on the second satellite remote sensing time series image of the breach area and the third satellite remote sensing time series image of the breach area to obtain a second image pixel difference value;

[0110] According to the first image pixel difference value and the second image pixel difference value, the satellite remote sensing image of the breach area is segmented into a change area to obtain a satellite remote sensing change area image of the breach area; the first image pixel difference value and the second image pixel difference value are analyzed for positive and negative changes to generate image pixel positive difference values ​​and image pixel negative difference values;

[0111] Based on the positive difference of image pixels, the satellite remote sensing change area image of the breach area is analyzed for terrain elevation changes to generate terrain elevation change data; based on the negative difference of image pixels, the satellite remote sensing change area image of the breach area is used to detect water body changes to generate water body change detection data; the terrain elevation change data and water body change detection data are used to visualize the macro-disaster risk of the breach area to generate a macro-disaster risk schematic diagram of the breach area.

[0112] In an embodiment of the present invention, by collecting satellite remote sensing images within a specified time period, it is ensured that the image time sequence is connected. Three frames of images are selected: the first frame (previous frame), the second frame (reference frame), and the third frame (next frame), which are named the first satellite remote sensing time sequence image of the breach area, the second satellite remote sensing time sequence image, and the third satellite remote sensing time sequence image, respectively. Image processing software (such as OpenCV or MATLAB) is used to perform pixel value difference calculation on the first satellite remote sensing time sequence image and the second satellite remote sensing time sequence image of the breach area to obtain the pixel difference value of the first image. Similarly, the pixel difference value of the second image is calculated, and the second satellite remote sensing time sequence image is differentiated from the third satellite remote sensing time sequence image. According to the pixel difference value of the first image and the pixel difference value of the second image, a threshold segmentation algorithm (such as the Otsu method) is applied to segment the change area to obtain a satellite remote sensing change area image of the breach area. The image segmentation result is visualized to highlight the change area. Perform positive and negative change analysis on the pixel difference value of the first image and the pixel difference value of the second image, extract the positive difference value (indicating the increased area) and negative difference value (indicating the reduced area) of the image pixels respectively, generate positive difference value and negative difference value data sets, and identify the nature of the change. Based on the positive difference value of the image pixel, perform terrain elevation change analysis on the satellite remote sensing change area image of the breach area, using the existing DEM data. Calculate the elevation change of the change area, generate terrain elevation change data, and output it as a visualization chart. Based on the positive difference value of the image pixel, perform terrain elevation change analysis on the satellite remote sensing change area image of the breach area, using the existing DEM data. Calculate the elevation change of the change area, generate terrain elevation change data, and output it as a visualization chart. Integrate the generated terrain elevation change data and water body change detection data, use GIS software (such as ArcGIS) for visualization processing, and generate a macro-disaster hazard diagram of the breach area, indicating potential dangerous areas and changes.

[0113] Preferably, performing flight path planning of the UAV according to the high-urgency node data and the low-urgency node data includes:

[0114] Perform node connection on high-urgency node data and low-urgency node data to generate a node connection network; perform priority path series connection on high-urgency node data through the node connection network to generate a high-urgency node flight path;

[0115] Confirm the node position of the high-urgency node flight path to obtain the high-urgency node position data; search for low-urgency nodes on the high-urgency node flight path through the high-urgency node position data to generate low-urgency node data with the same position;

[0116] The same-direction low-urgency node data and high-urgency node flight paths are merged into the same-direction paths to generate the same-direction node flight paths; the node connection network is segmented according to the same-direction node flight paths to generate the same-direction UAV flight paths; the adjacent channel angles are calculated for the same-direction UAV flight paths to obtain the adjacent channel angle data;

[0117] The flight course of the UAV with the same azimuth is transformed into the flight course with the minimum angle through the angle data of adjacent courses, and the flight course planning path data of the UAV is generated.

[0118] In the embodiment of the present invention, by collecting high-urgency node data and low-urgency node data, the data format is ensured to be consistent. Using the graph structure in graph theory, the nodes are used as vertices in the graph, and the edges between high-urgency nodes and low-urgency nodes are connected to generate a node connection network. In the node connection network, the priority strategy is used to plan the path of the high-urgency node to ensure the shortest path, generate the high-urgency node flight path, and record all nodes and edges in the path. The azimuth of each node in the high-urgency node flight path is calculated using the following formula: in and is the ordinate of the drone flight at the next timestamp and the abscissa of the drone flight at the next timestamp, and The ordinate of the drone flight at the previous timestamp and the abscissa of the drone flight at the next timestamp are used to obtain the high-urgency node position data and identify the flight direction of each node. Based on the high-urgency node position data, search for low-urgency nodes with the same position in the node connection network. Record all qualified low-urgency nodes and generate low-urgency node data with the same position. Merge the low-urgency node data with the flight path of the high-urgency node to ensure the continuity and feasibility of the new path, and generate the flight path of the node with the same position, including all relevant nodes. According to the flight path of the node with the same position, segment the node connection network and delineate the drone flight path. Ensure that the segmented channel has sufficient width and safety spacing. Calculate the angle between adjacent channels using the following formula: in Represented as the direction vector of the first channel, Expressed as the direction vector of another channel, the adjacent channel angle data is obtained and the angle information is recorded. According to the adjacent channel angle data, the minimum angle is analyzed and determined to optimize the flight path. The flight channel of the UAV in the same direction is adjusted to meet the minimum angle requirement, and the final UAV flight channel planning path data is generated.

[0119] Preferably, analyzing the damage degree of the dam using the microscopic high-precision breach image includes:

[0120] The breach area contour is segmented on the microscopic high-precision breach image to generate the breach area contour image; the number of pixels in the damaged area is calculated on the breach area contour image to obtain the number of pixels in the damaged area; the number of pixels in the damaged area and the microscopic high-precision breach image are converted into actual geographic size ratios to generate the dam damaged area data;

[0121] Perform multi-view image screening on microscopic high-precision breach images to generate a multi-view image set of microscopic breaches; perform feature point extraction on the multi-view image set of microscopic breaches to generate multi-view image feature points; perform feature point matching on the multi-view image feature points to generate multi-view matching feature points;

[0122] Parallax calculation is performed on multi-view matching feature points to obtain the multi-view parallax calculation results of the breach; binocular reconstruction is performed on the multi-view parallax calculation results of the breach to generate a dam damage depth map; depth estimation is performed on the dam damage depth map to generate dam damage depth data; the dam damage area data and dam damage depth data are visualized to generate a macro-disaster hazard diagram of the breach area.

[0123] In an embodiment of the present invention, a denoising process is performed on a microscopic high-precision breach image, and a Gaussian filter or a median filter is used to reduce image noise. An edge detection algorithm (such as Canny edge detection) is applied to detect the breach edge. A contour extraction algorithm (such as Hough transform or image watershed algorithm) is used to generate a contour image of the breach area. A black and white image is output, in which the white part represents the contour of the breach area and the black part represents the background. The breach area contour image is analyzed, the number of white pixels is calculated, and the number of pixels in the damaged area is obtained. According to the known image size and the actual geographic size ratio, the number of pixels in the damaged area is converted into the actual dam damaged area data. Use the formula: in is the number of pixels in the damaged area, For actual size, is the number of image pixels. Multi-view images related to microscopic high-precision breaches are selected from images collected at different angles to ensure that each view covers the breach area. Multi-view images are organized into sets for subsequent feature point extraction. Feature point detection algorithms (such as SIFT, SURF, or ORB) are used to extract feature points in multi-view images. Feature matching algorithms (such as FLANN or brute force matching) are used to match feature points from different viewpoints to generate a multi-view matching feature point set. Based on the matched feature points, the disparity value is calculated using the formula: disparity = left image coordinates − right image coordinates. Based on the disparity information, binocular stereo vision reconstruction technology is used to generate a dam damage depth map. Triangulation is used to calculate the depth information through disparity and camera internal parameters: depth = f⋅R / disparity, where f is the focal length of the camera and R is the baseline distance between the two cameras. The depth of the generated dam damage depth map is estimated, and areas with larger depth values ​​are identified as severely damaged areas. The dam damage area data and dam damage depth data are visualized, and the macro-disaster hazard diagram of the breach area is displayed using heat maps or 3D graphics to generate the final macro-disaster hazard diagram of the breach area, which is convenient for relevant personnel to make decisions and respond.

[0124] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0125] Step S31: extracting the breach space features from the breach rapid situation awareness image to obtain a breach space feature matrix, wherein the breach space feature extraction includes texture feature extraction and shape feature extraction;

[0126] Step S32: dividing the breach space feature matrix into a data set to obtain a model training set and a model test set; training the model training set using a long short-term memory neural network algorithm to generate a breach situation prediction pre-model; optimizing and iterating the breach situation prediction pre-model using the model test set to generate a breach situation prediction model;

[0127] Step S33: importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction map;

[0128] Step S34: color-code the breach situation prediction map to generate a breach situation prediction coded map.

[0129] In an embodiment of the present invention, a Gaussian filter or a median filter is used to perform denoising on the rapid situational awareness image of the breach. Texture features are extracted using a gray level co-occurrence matrix (GLCM), including indicators such as contrast, entropy, uniformity, and correlation. The eigenvalues ​​of the GLCM are calculated, a texture feature vector is constructed, and a texture feature matrix containing various texture features is generated. The contour of the breach area is extracted using an edge detection algorithm (such as Canny edge detection). The shape features are calculated using Hu invariant moments (Hu Moments) or shape descriptors (such as rectangularity and roundness). The extracted shape feature vector should include area, perimeter, aspect ratio, etc. The shape features are integrated into the shape feature matrix. The texture feature matrix and the shape feature matrix are merged to form the final breach space feature matrix. The breach space feature matrix is ​​randomly divided into a training set (e.g., 70%) and a test set (30%) to ensure that the samples are evenly distributed. An LSTM model is constructed, the input layer, the LSTM layer, and the output layer are set, and a loss function (such as mean square error) and an optimizer (such as Adam) are defined. Use the model training set to train the LSTM model, iterate multiple cycles (epochs), and monitor the training loss and validation loss in real time. After the training is completed, save the trained model as the "Break Situation Prediction Prediction Model". Use the test set to evaluate the "Break Situation Prediction Prediction Model" and calculate performance indicators (such as accuracy and recall). Adjust the model hyperparameters (such as learning rate and batch size) according to the test results and retrain until satisfactory performance is achieved. Save the optimized "Break Situation Prediction Model". Import the new breach rapid situation awareness image data into the "Break Situation Prediction Model". Use the model to predict the spatiotemporal evolution and output the breach situation prediction map. Define a color mapping table based on the value range in the breach situation prediction map (for example: low-risk areas are green, medium-risk areas are yellow, and high-risk areas are red). Convert the values ​​in the prediction map to corresponding color values ​​to generate a breach situation prediction coding map. Finally, a visual breach situation prediction coding map is generated to facilitate quick judgment of risk areas.

[0130] In this specification, a breach rapid situation awareness system based on multi-source data is provided, which is used to execute the above-mentioned breach rapid situation awareness method based on multi-source data. The breach rapid situation awareness system based on multi-source data includes:

[0131] The breach sensing module is used to obtain the data of the dam monitoring station; to set up the hydrological sensor network for the dam monitoring station data to obtain the dam hydrological sensor network data; to perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; to perform breach node sensing on the dam hydrological sensor network data based on the ultrasonic water level change trend graph to generate breach node data and breach prediction node data.

[0132] The breach image analysis module is used to perform macro-hazard risk analysis of the breach area based on breach node data and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV flight route planning path data and generate a micro-hazard diagram of dam damage; perform image integration of the macro-hazard diagram of the breach area and the micro-damage diagram of the dam area to generate a rapid situation awareness image of the breach;

[0133] The breach dynamic evolution module is used to extract breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; construct a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; import the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map;

[0134] The breach perception and warning module is used to set the situation warning threshold for the breach situation prediction coding map to obtain the situation warning threshold; perform situation warning on the breach rapid situation perception image according to the situation warning threshold to generate a situation warning map; perform multi-party collaborative situation perception response on the situation warning map to generate a visual breach situation perception shared atlas.

[0135] The beneficial effect of the present invention is that by acquiring the data of the dam monitoring station and setting up the hydrological sensor network, the collected data can be ensured to be accurate and comprehensive, laying the foundation for subsequent analysis. Through ultrasonic water level change trend analysis, water level changes can be monitored in real time, abnormal conditions can be discovered in time, and it is helpful to respond quickly to potential risks. The breach node data and prediction node data generated according to the water level change trend are helpful to identify and predict breach risks and improve the predictability of emergency management. The macro-disaster risk analysis of the breach area and the generated danger diagram can provide decision makers with a clear risk view and help evaluate potential impacts. The UAV flight route planning based on the breach node data can optimize the UAV monitoring path, improve monitoring efficiency, and ensure that key areas receive focus. The micro-dam damage degree analysis can provide a detailed damage assessment, help decision makers understand the actual damage situation, and formulate corresponding repair plans. The macro and micro damage diagrams are image-integrated, and the generated rapid situation awareness image provides decision makers with an intuitive perspective and helps to quickly judge the current situation. The feature matrix obtained by spatial feature extraction can reveal the spatial distribution and characteristics of the breach and provide rich information for subsequent model construction. The situation prediction model constructed based on the feature matrix can make scientific predictions using historical data and current conditions, and enhance the foresight of disaster management. By importing the rapid situation awareness image into the prediction model for spatiotemporal evolution, the current situation can be dynamically updated to help decision makers adjust response strategies in a timely manner. By setting the situation warning threshold, effective monitoring of potential dangers can be achieved to ensure that appropriate measures are taken before the danger occurs. The generation of situation warning maps promotes the collaboration of all parties, ensures timely communication and collaboration between relevant departments, and improves the efficiency of emergency response. The generation of a shared atlas of visual breach situation awareness promotes the transparency and sharing of information, and provides strong decision-making support for managers at all levels. Therefore, the present invention improves the accuracy and response speed of rapid situation awareness of breaches through real-time data integration, comprehensive risk assessment, drone monitoring, dynamic prediction models, and multi-party collaborative response.

[0136] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0137] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for rapid breach situation awareness based on multi-source data, characterized in that: The following steps are involved: Step S1: Acquire dam monitoring site data; The hydrological sensor network is set up for the dam monitoring station data to obtain the dam hydrological sensor network data; the ultrasonic water level change trend analysis is performed on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; the breach node perception is performed on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data; Step S2: Perform macro-hazard risk analysis of the breach area according to breach node data, and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data, and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV according to the UAV flight route planning path data, and generate a micro-disaster diagram of dam damage; perform image integration of the macro-disaster diagram of the breach area and the micro-damage diagram of the dam area, and generate a rapid situation awareness image of the breach; Step S3: extracting the breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; constructing a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map; Step S4: setting a situation warning threshold for the breach situation prediction coding map to obtain a situation warning threshold; According to the situation warning threshold, situation warning is carried out on the rapid situation awareness image of the breach to generate a situation warning map; multi-party collaborative situation awareness response is carried out on the situation warning map to generate a visual breach situation awareness shared atlas.

2. The method for rapid breach situation awareness based on multi-source data according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Acquire dam monitoring site data; Step S12: performing hydrological sensor network setting on the dam monitoring station data to obtain dam hydrological sensor network data, wherein the dam hydrological sensor network includes an ultrasonic water level meter and a pressure sensor; performing ultrasonic water level measurement data collection on the dam hydrological sensor network data to obtain ultrasonic water level collection data; Step S13: preprocessing the ultrasonic water level acquisition data to generate standard ultrasonic water level acquisition data, wherein the data preprocessing includes data cleaning, data missing value filling and data standardization; Step S14: Analyze the water level change trend of the standard ultrasonic water level acquisition data to generate an ultrasonic water level change trend graph; perform breach node perception on the dam hydrological sensor network data based on the ultrasonic water level change trend graph based on the pressure sensor to generate breach node data and breach prediction node data.

3. The method for rapid breach situation awareness based on multi-source data according to claim 2 is characterized in that: Step S14 includes the following steps: Step S141: performing timestamp and water level extraction on the standard ultrasonic water level acquisition data to obtain water level acquisition timestamp and water level data; using the water level acquisition timestamp as the x-axis and the water level data as the y-axis to perform time series discrete point mapping, thereby generating a water level change line graph; Step S142: Calculate the water level change rate of the water level change line graph to generate water level change rate data; segment the water level change line graph into fluctuation change line segments according to the water level change rate data to generate an ultrasonic water level change trend graph, wherein the ultrasonic water level change trend graph includes fluctuation change line segments and gentle change line segments; perform abnormal fluctuation screening on the fluctuation change line segments based on a preset change rate threshold to obtain abnormal fluctuation line segment data; Step S143: Perform trend line fitting on the gently changing line segments to generate gently changing trend line segments; perform abnormal fluctuation screening on the fluctuating line segments based on a preset change rate threshold to obtain abnormal fluctuation fitting line segment data; perform breach prediction area division on the dam hydrological sensor network data according to the gently changing trend line segments to generate breach first prediction area data; perform suspicious breach area division on the dam hydrological sensor network data according to the abnormal fluctuation line segment data to generate suspicious breach area data; Step S144: Based on the pressure sensor, the dam area pressure analysis is performed on the suspected breach area data and the breach first prediction area data to generate the dam area pressure data; the breach node perception is performed on the dam hydrological sensor network data through the dam area pressure data to generate the breach node data and the breach prediction node data.

4. The method for rapid breach situation awareness based on multi-source data according to claim 2 is characterized in that: The breach node perception of the dam hydrological sensor network data through the dam area pressure data includes: The pressure data of the dam area is used to detect the pressure. When the pressure data of the dam area is detected to be empty, the breach node of the dam hydrological sensor network data is marked to generate breach node data. When the dam area pressure data detection is not empty, the dam area pressure data is screened for the dam double-side pressure to generate the first lateral pressure data and the second lateral pressure data; the first lateral pressure data and the second lateral pressure data are calculated for the pressure balance difference to obtain the dam pressure lateral difference data; The suspected breach orientation is estimated through the lateral difference data of the dam pressure to obtain the breach second prediction area data; the breach nodes of the dam hydrological sensor network data are marked according to the breach first prediction area data and the breach second prediction area data to generate breach prediction node data.

5. The method for rapid breach situation awareness based on multi-source data according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: collecting satellite remote sensing images of the breach area according to breach node data to generate satellite remote sensing images of the breach area; performing macro-hazard risk analysis of the breach area on the satellite remote sensing images of the breach area to generate a macro-hazard risk diagram of the breach area; Step S22: sorting the breach node data and the breach prediction node data by urgency, marking the breach node data as high-urgency node data, and marking the breach prediction node data as low-urgency node data; performing UAV flight route planning based on the high-urgency node data and the low-urgency node data, and generating UAV flight route planning path data; Step S23: taking microscopic high-precision breach photos according to the UAV flight route planning path data, thereby generating a microscopic high-precision breach image; analyzing the degree of dam damage on the microscopic high-precision breach image, and generating a microscopic disaster diagram of dam damage; Step S24: integrating the macroscopic disaster diagram of the breach area and the microscopic damage diagram of the dam area to generate a rapid situation awareness image of the breach.

6. The method for rapid breach situation awareness based on multi-source data according to claim 5 is characterized in that: The macro-hazard analysis of the breach area based on satellite remote sensing images of the breach area includes: Performing connected time-series image screening on the satellite remote sensing images of the breach area to obtain a first satellite remote sensing time-series image of the breach area, a second satellite remote sensing time-series image of the breach area, and a third satellite remote sensing time-series image of the breach area, wherein the first satellite remote sensing time-series image of the breach area is a previous frame image of the second satellite remote sensing time-series image of the breach area, and the third satellite remote sensing time-series image of the breach area is a subsequent frame image of the second satellite remote sensing time-series image of the breach area; Performing image pixel value difference calculation on the first satellite remote sensing time series image of the breach area and the second satellite remote sensing time series image of the breach area to obtain a first image pixel difference value; performing image pixel value difference calculation on the second satellite remote sensing time series image of the breach area and the third satellite remote sensing time series image of the breach area to obtain a second image pixel difference value; According to the first image pixel difference value and the second image pixel difference value, the satellite remote sensing image of the breach area is segmented into a change area to obtain a satellite remote sensing change area image of the breach area; the first image pixel difference value and the second image pixel difference value are analyzed for positive and negative changes to generate image pixel positive difference values ​​and image pixel negative difference values; Based on the positive difference of image pixels, the satellite remote sensing change area image of the breach area is analyzed for terrain elevation changes to generate terrain elevation change data; based on the negative difference of image pixels, the satellite remote sensing change area image of the breach area is used to detect water body changes to generate water body change detection data; the terrain elevation change data and water body change detection data are used to visualize the macro-disaster risk of the breach area to generate a macro-disaster risk schematic diagram of the breach area.

7. The method for rapid breach situation awareness based on multi-source data according to claim 5 is characterized in that: The flight path planning of UAV based on high-urgency node data and low-urgency node data includes: Perform node connection on high-urgency node data and low-urgency node data to generate a node connection network; perform priority path series connection on high-urgency node data through the node connection network to generate a high-urgency node flight path; Confirm the node position of the high-urgency node flight path to obtain the high-urgency node position data; search for low-urgency nodes on the high-urgency node flight path through the high-urgency node position data to generate low-urgency node data with the same position; The same-direction low-urgency node data and high-urgency node flight paths are merged into the same-direction paths to generate the same-direction node flight paths; the node connection network is segmented according to the same-direction node flight paths to generate the same-direction UAV flight paths; the adjacent channel angles are calculated for the same-direction UAV flight paths to obtain the adjacent channel angle data; The flight course of the UAV with the same azimuth is transformed into the flight course with the minimum angle through the angle data of adjacent courses, and the flight course planning path data of the UAV is generated.

8. The method for rapid breach situation awareness based on multi-source data according to claim 5 is characterized in that: The damage analysis of dams based on microscopic high-precision breach images includes: The breach area contour is segmented on the microscopic high-precision breach image to generate the breach area contour image; the number of pixels in the damaged area is calculated on the breach area contour image to obtain the number of pixels in the damaged area; the number of pixels in the damaged area and the microscopic high-precision breach image are converted into actual geographic size ratios to generate the dam damaged area data; Perform multi-view image screening on microscopic high-precision breach images to generate a multi-view image set of microscopic breaches; perform feature point extraction on the multi-view image set of microscopic breaches to generate multi-view image feature points; perform feature point matching on the multi-view image feature points to generate multi-view matching feature points; Parallax calculation is performed on multi-view matching feature points to obtain the multi-view parallax calculation results of the breach; binocular reconstruction is performed on the multi-view parallax calculation results of the breach to generate a dam damage depth map; depth estimation is performed on the dam damage depth map to generate dam damage depth data; the dam damage area data and dam damage depth data are visualized to generate a macro-disaster hazard diagram of the breach area.

9. The method for rapid breach situation awareness based on multi-source data according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: extracting the breach space features from the breach rapid situation awareness image to obtain a breach space feature matrix, wherein the breach space feature extraction includes texture feature extraction and shape feature extraction; Step S32: dividing the breach space feature matrix into a data set to obtain a model training set and a model test set; training the model training set using a long short-term memory neural network algorithm to generate a breach situation prediction pre-model; optimizing and iterating the breach situation prediction pre-model using the model test set to generate a breach situation prediction model; Step S33: importing the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction map; Step S34: color-code the breach situation prediction map to generate a breach situation prediction coded map.

10. A breach rapid situation awareness system based on multi-source data, characterized in that: The method for rapid situational awareness of breach based on multi-source data according to claim 1 is used to execute the method for rapid situational awareness of breach based on multi-source data, and the rapid situational awareness system for breach based on multi-source data comprises: The breach sensing module is used to obtain the data of the dam monitoring station; set up the hydrological sensor network for the dam monitoring station data to obtain the dam hydrological sensor network data; perform ultrasonic water level change trend analysis on the dam hydrological sensor network data to generate an ultrasonic water level change trend graph; perform breach node sensing on the dam hydrological sensor network data according to the ultrasonic water level change trend graph to generate breach node data and breach prediction node data; The breach image analysis module is used to perform macro-hazard risk analysis of the breach area based on breach node data and generate a macro-hazard risk diagram of the breach area; perform UAV flight route planning on breach node data and breach prediction node data and generate UAV flight route planning path data; perform micro-damage degree analysis on the UAV flight route planning path data and generate a micro-hazard diagram of dam damage; perform image integration of the macro-hazard diagram of the breach area and the micro-damage diagram of the dam area to generate a rapid situation awareness image of the breach; The breach dynamic evolution module is used to extract breach spatial features from the breach rapid situation awareness image to obtain a breach spatial feature matrix; construct a situation prediction model based on the breach spatial feature matrix to generate a breach situation prediction model; import the breach rapid situation awareness image into the breach situation prediction model for spatiotemporal evolution to generate a breach situation prediction coding map; The breach perception and warning module is used to set the situation warning threshold for the breach situation prediction coding map to obtain the situation warning threshold; perform situation warning on the breach rapid situation perception image according to the situation warning threshold to generate a situation warning map; perform multi-party collaborative situation perception response on the situation warning map to generate a visual breach situation perception shared atlas.

Images